Horse_ebooks Is Human After All | Hacker News Reader

Horse_ebooks Is Human After All

newyorker.com

100 pointsdominicgs31 comments

Am I alone in being slightly disappointed to discover that Horse_ebooks is not a robot?

I've had to conduct a mini-self analysis on this news, as I find myself disappointed. Why?

All I can come up with is that I found joy in the befuddlement caused by my brain trying to make sense of the non-sensicle almost-marketing speak. The fact that it was (supposedly) accidental somehow added to the joy. But I can't think of a good reason why. Was it that I thought some spammer was wasting clock-cycles making me laugh rather than selling me viagra?

It reminded me of Steven Frank's "Spamusement!" (which I also found very funny back in the days of email.)

http://spamusement.com

And Panic's love of Horse_ebooks is where I first learned of it.

EDIT: I've just been back to spamusement.com and, yes, it is still really funny. DO NOT start looking unless you have half an hour to waste!

I'm relieved. It never occurred to me that horse_ebooks was a spam bot until I started reading other people stating with certainty that it is a spambot. It always struck me as a human posing as a spambot because it was too clever and poetic.

It also surprises me that so many people think we're more likely to encounter a legitimately clever spam bot than a clever human impersonating a spambot.

I agree. The phrases consistently end in a way that is never completely resolved. See:

"In less than three weeks you can be looking at things differently, more creatively, if you"

"Why many people achieve very little despite spending most of their time"

"The interviews can be shorter since you would have many"

I only ever saw retweeted or starred tweets from it (as I quit direct twitter a while back, deciding instead to view it only through Stellar).

I had assumed that what I saw was a tiny proportion of tweets — those that were funny and/or poetic. If I'd know they all were than I'm sure I would have never thought it was a bot.

> It always struck me as a human posing as a spambot because it was too clever and poetic

I dunno, I've seen some pretty clever and poetic quotes from my own markov-chain powered IRC bot:

http://quotes.shishnet.org/quotes?search=MrCuddlemews

(#cuddles on darkmyst to poke him yourself; hilight by name or by starting a line with "@")

And the inspiration: http://www.jibble.org/montyquotes.php

I didn't expect it to be a spambot per se, but I did expect it to be some sort of automated scraping of the site. Because "human posts semi-random words" is more or less trivial, but "human teaches neural net to perform broken poetry while scraping an ebooks site" is much more exciting.

I wanted it to be something like http://thedailywtf.com/Comments/No,_We_Need_a_Neural_Network...

If anyone is interested in running an _ebooks account of their own, Jacob Harris of the New York Times wrote a Ruby library called iron_ebooks[1] which does just that. Using iron_ebooks, Harris has two accounts: @harrisj_ebooks, which uses his main twitter handle (@harrisj) as an input, and @harrisj_ebooks2 which uses @harrisj_ebooks as its input. The generated tweets have actually been pretty funny and well done.

Also, iron_ebooks uses Iron.io, so you get all the fun of an _ebooks account with all the "serverless" greatness of IronWorker.

[1] https://github.com/harrisj/iron_ebooks

@Horse_ebooks is (was?) an excellent piece of performance art. I think the modern audience still has a sense of wonderment when it comes to the computer-generated (as though it is the computer itself doing the generating and not really the programmer who wrote the algorithm). People were willing to believe that "computers" were at the point where they could create consistently hilarious phrases, and are disappointed (and even outraged) today when they find that we're not there yet. The artists have imparted the audience with a sense of disillusionment. Perhaps we'll see more art in this style in the future, although part of (or all of) the magic seems to be in the audience's belief that there was no human intelligence behind the content, so perhaps this isn't duplicatable. If only I could set this experience to a Kraftwerk soundtrack, and maybe play Bladerunner on a screen in the corner.

I can't wait to see what this group comes up with next!

The NY Times has been running an automated haiku experiment for some time now. I think it has a lot to do with your comment. Check this: haiku.nytimes.com

Ha! That's great, thanks! I had a feeling Jacob Harris was behind that, as he was also behind @nytimes_ebooks.

http://en.wikipedia.org/wiki/Racter was another case like that (apparently; I haven't followed horse_ebooks but I was gulled by Racter at the time: I found it hard to believe but I wasn't cynical enough to disbelieve it).

Horse_ebooks was the inspiration for other novelty Twitter accounts such as horse_js ( https://twitter.com/horse_js ) which is somebody taking quotes related to JS out of context, and horse_esdiscuss.

As an iOS developer, I always enjoy @Horse_iOS.

Gawker has a writeup of the event at http://gawker.com/horse_ebooks-has-been-a-buzzfeed-employee-...

The exhibit is three people sitting at a table answering phones (specifically not interacting with the gallery) along with formalistic artistic ambiance.

Sidenote: You've got to give them credit. This has been their endgame for years.

Horse_ebooks, meet Pet Rock. Pet Rock, Horse_ebooks.

For those interested, here's some background on the pronunciation book side of the story:

http://77days.wikia.com/wiki/Pronunciation_Book_Conspiracy_W...

Thank you, this was an interesting read.

These guys are good and definitely know their audience.

What I'm still wondering... was Horse_ebooks was linking to completely unaffiliated spammy content, or did they build an entire constellation of shitty websites?

And, most importantly, who actually wrote this visitor-triggered cron job script[1][2]?

[1] http://quotestatusjoke.com/twitter/tw.php

[2] https://news.ycombinator.com/item?id=6171355

EDIT: A ha! It appears they purchased Horse_ebooks from an actual Russian web developer in 2011 (who ostensibly actually built quotestatusjoke.com), and linked to legitimate crappy third party self-help ebook websites.

A gallery installation where you can watch the artists read tweets to callers on the phone seems like an odd and unimaginative way to end it, but maybe I just don't get it.

My favorite mystery blog: http://moneyistheway.blogspot.com/

He's the world's foremost financial shaman. The blog ranges from the bizarre to the illucid.

Another awesome one along a slightly different genre line: http://www.johntitor.com/

https://news.ycombinator.com/item?id=4697400

You're in for a treat.

To be totally honest, it wasn't until I saw the pronunciation guide that I realized it might be called "horsey books" instead.

I used to work with people who called it horsey books. I thought they were morons for doing that. Turns out they were morons for completely different reasons now?

Be sure to drink your Ovaltine!

So this was an account that posted random Tweets, and it somehow has 200,000 followers? Why? What am I missing?

Internet novelty fad syndrome growth by exponential ego expansion?

"I've found this thing I think is cool! I want other people to like me, so I'll tell them about it. If they like it, they'll thank me for showing it to them and my life will have meaning!"

Because it was funny, and it was interesting to see what an entirely random algorithm would turn up.

Of course, it wasn't an algorithm but it's not like the people following knew that.

According to Gawker, the current owners purchased the account from a Russian spammer in 2011. Before that it presumably was indeed an algorithm for promoting crappy ebooks and not an art project.

Yes, I think it's been known for some time that it "changed" in 2011 (it stopped being posted "from web", IIRC). Still, an interesting story.

Agreed. Another interesting thing I've seen some creative users do is RT various selected @horse_ebooks tweets in succession to make complete statements or sort-of "@horse_ebooks poems" When you saw one done well, it was pretty clever/funny.

I think a lot of people always suspected it was a person(s), but it was still interesting in the sense of wondering and imagining a person emulating a bot emulating a person.

My favorite tweet:

> Unfortunately, as you already know, people.

You might also just not get it. Everyone's tastes are different.

Reply on news.ycombinator.com